The structure of interfaces probed by sum‐frequency spectroscopy

Abstract

The chemistry of interfaces plays a fundamental role in adhesion. Many analytical methods, such as SIMS, XPS, AES and SEM, can be used to study surfaces after adhesive failure. Only a few techniques exist, however, for the non-destructive, in situ chemical analysis of buried interfaces. Infrared-visible sum-frequency generation (SFG) is one such technique. Shen has demonstrated recently' that infrared spectra of interfaces can be obtained by sum-frequency spectroscopy, even in the presence of bulk phases on either side. SFG is a powerful technique still in its infancy; here, we demonstrate the application of SFG to the analysis of metals coated with monolayer films. These films are useful model systems and also have practical importance as adhesion promoters and wetting modifiers.In infrared-visible sum-frequency generation, a fixedfrequency visible laser and a tunable infrared laser are overlapped at an interface and the light emitted at the sum of the incident frequencies, u , , , , is detected (Scheme 1). In our experiments, laser light at 532 nm is provided by a frequencydoubled Nd : YAG laser, and the tunable infrared radiation is generated by difference frequency mixing of the Nd : YAG fundamental and a near-IR, Nd:YAG-pumped dye laser. The photons emitted at the sum frequency are separated from scattered light by a filter and monochromator, and detected with a photomultiplier tube. The intensity, S,,, , of the emitted sum frequency is given by' (photons per pulse) where Li are Fresnel field factors, I is the intensity of a laser, A the area of overlap of the laser beams, T the pulse duration and $, ) , , ( -us,,; wvis, urn) the second-order non-linear susceptibility of the sample.

Dynamic Duo: Vibrational Sum Frequency Scattering Investigation of pH-Switchable Carboxylic Acid/Carboxylate Surfactants on Nanodroplet Surfaces

“…The resonant contribution contains information about the orientation and population of the resonant vibrational dipole moment probed. The spectra were fit by using a procedure that follows previous Richmond group analysis, 49 based on a procedure described by Bain et al 50 This describes the origin of the χ (2) response, given as a convolution of a Lorentzian and Gaussian distribution:…”

“…The resonant contribution contains information about the orientation and population of the resonant vibrational dipole moment probed. The spectra were fit by using a procedure that follows previous Richmond group analysis, based on a procedure described by Bain et al This describes the origin of the χ (2) response, given as a convolution of a Lorentzian and Gaussian distribution: Equation takes into account the homogeneous broadening (Γ L ) due to the natural decay of the vibrational mode probed and the inhomogeneous broadening (Γ v ) arising from the molecular environment of the mode. Each peak in the spectra was fit by using five parameters: amplitude ( A v ), phase (ψ v ), Lorentzian line width (Γ L ), peak position frequency (ω v ), and Gaussian line width (Γ v ).…”

Coming to Order: Adsorption and Structure of Nonionic Polymer at the Oil/Water Interface as Influenced by Cationic and Anionic Surfactants

“…Fits to spectra follow previous Richmond publications, 51 where the procedure described by Bain et al 52 fits line shapes to a convolution of a Gaussian and Lorentzian distribution (eq 1), taking into account the homogeneous broadening (Γ L ) caused by the nature of the vibrational transition as well as the inhomogeneous broadening (Γ v ) caused by the molecular environment of the detected vibrational modes. Each peak is fit to five parameters: amplitude (A v ), phase (φ v ), Lorentzian line width (Γ L ), frequency (ω v ), and Gaussian line width (Γ v ).…”

Molecular Interactions Leading to the Coadsorption of Surfactant Dodecyltrimethylammonium Bromide and Poly(styrenesulfonate) at the Oil/Water Interface

“…All spectra were taken in D 2 O and are the average of three VSF spectra taken on separate days. When appropriate, spectra were fit to a convolution of a Gaussian and a Lorentzian distribution originally described by Bain et al 49 The fitting equation, all fitting parameters, and peak assignments are shown in equation S1 and Tables S2 − S4.…”